O.R. study faults reliance on click-through rates to assess banner ads

Viewing of internet ads does lead to future sales

Contrary to popular e-wisdom, measuring Internet banner ads only by the number of times that viewers click through is faulty, according to a paper presented today at a conference of the Institute for Operations Research and the Management Sciences (INFORMS®).

The paper contends that the more often consumers see online banner ads, the more likely they are to make a purchase, even if they don’t click through immediately to the advertiser’s internet storefront. The evidence also suggests that banner advertising may be a useful tactic for managing relationships with repeat-customers.

“We are finding strong evidence of exposure effects,” explains the University of Chicago’s Jean-Pierre Dubé, one of the authors. “In other words, even if a consumer does not click through on the banner, they still see it and, thus, there is an impression.

“For existing repeat-consumers, we find the time between visits to a site and purchasing to be affected by the accumulated impressions, even if there is no direct click-through. We also find that accumulated impressions increase the total expenditure at the time of purchase. Moreover, these impressions have lasting effects on consumers. Interestingly, we also find that inter-purchase times are lower if consumers are exposed to the firm’s advertising on many different websites.”

The research adds to the controversy over the rewards of advertising on the Internet. Companies concerned about the value of advertising online have sought more concrete ways of determining the ability of online ads to spur sales. Although click-through rates don’t measure actual purchases, they have been viewed as a useful indicator of customer interest. The new study brings the discussion back to the question of whether latent effects of online advertising are significant and measurable.

The paper, “The Effects of Banner Advertising on Consumer Inter-Purchase Times and Expenditures in Digital Environments,” is by Dr. Dubé, Puneet Manchanda, Khim Yong Goh, and Pradeep K. Chintagunta of the University of Chicago Graduate School of Business. It is being delivered today at the 2002 Marketing Science Conference, sponsored by the INFORMS College on Marketing. The conference is running from June 27-30 at the University of Alberta.

Data

The data come from an anonymous Internet-only firm engaged in selling healthcare and beauty products, as well as non-prescription drugs, to consumers. The data span a period of three months during the third quarter of 2000. The data are available at the individual cookie level. The data are unique in that they survey individual-level stimulus (advertising) and response (purchase amount).

The data are contained in three databases. One database comprises the on-line advertisement banner exposure and click-through response originating from promotional campaigns that were run on websites such as Yahoo!, AOL, Women.com, iVillage.com, Healthcentral.com, and E*Trade. A second database contains the date and time of the purchase transaction for each unique cookie identifier. The third database consist of the total dollar value of the transaction and a key that links these values with the cookie identifier in the first database.

Analyzing online repeat customers

The paper focuses on repeat visit and purchase behavior of consumers at a website. Among the authors’ findings are:

• Click-through is a poor measure of advertising effectiveness because it accounts for a very small proportion of overall purchases.

• Examining the effects of advertising on consumers, the authors find that the time since last exposure and the number of creatives have a much larger effect on purchase timing relative to the number of exposures.

This result has implications for advertising copy and ad timing. In terms of copy, the authors conclude that exposing the same consumer to several unrelated creatives may be less beneficial than a single creative.

In terms of timing, they speculate that it may be more beneficial to expose consumers to a series of evenly spaced ads relative to massed exposures. Since there is a strong same-day purchase effect given exposure, the authors suggest that advertisers use banner ads to smooth out sales and run special promotions.

• Findings about when shoppers make an online purchase imply that managers should ensure that once consumers reach the site their shopping baskets are as large as possible in dollar terms. Thus, the authors suggest, for example, following the lead of Internet marketers who offer “point of checkout” promotions to increase basket expenditures.

• There is a strong positive benefit of exposing customers to the same ad across many websites.

The Institute for Operations Research and the Management Sciences (INFORMS®) is an international scientific society with over 10,000 members dedicated to applying scientific methods to help improve decision-making, management, and operations. Members of INFORMS work in business, government, and academia. They are represented in fields as diverse as airlines, health care, law enforcement, the military, the stock market, and telecommunications. 2002 is the 50th anniversary of organized operations research in the United States. 1952 was the year that the journal Operations Research and the Operations Research Society of America, one of the founding societies of INFORMS, were born. The INFORMS website is at http://www.informs.org.